24-hour unmanned autonomous excavator emerges as 'legal' infinite labor machine

In a former quarry in Switzerland, an HD Hyundai Construction Equipment's 22-ton excavator, 'Devon DX225LC,' autonomously dug a 300-meter-long ditch for landslide prevention. The machine, equipped with an AI-based autonomous solution called 'Real-X,' scanned the terrain, determined the optimal path, and executed the digging with precision, impressing local construction firm Kibag. This demonstration highlights the potential of autonomous heavy machinery in addressing the growing shortage of skilled labor in the construction sector.

“

Excavators perform various tasks. In drainage work, they don't just dig; they move pipes and break rocks. We are developing autonomous operations for these tasks one by one. Currently, we are focusing on developing digging and dumping technologies.

HD Hyundai Construction Equipment claims that Real-X can increase productivity by up to 120% compared to human operators, as it is free from fatigue and concentration lapses. The system uses AI to analyze terrain data from cameras and LiDAR, precisely controlling the excavator's arm and bucket to follow optimal paths. Jang Gye-bong, a senior researcher involved in Real-X's development, explained that the system is being trained to handle various tasks beyond simple digging, including moving pipes and breaking rocks, with current focus on digging and dumping operations.

“

Autonomous driving equipment that increases the efficiency of excavation and loading work is an attractive investment target in a situation where the shortage of skilled labor is worsening.

The development of autonomous construction equipment is gaining momentum globally. At the CES 2024, Caterpillar showcased autonomous trucks operating 24/7 in quarries and mines, and also presented excavators, bulldozers, and loaders with autonomous capabilities, though full commercialization for complex tasks like excavation is still pending.

“

Because it is free from the increase in fatigue and decrease in concentration that human operators inevitably experience, it can achieve 120% productivity compared to manned operation.

Real-X undergoes training in both physical and virtual environments, learning from digitized operational data of skilled human workers. Sensors on the excavator's components capture control signals and operational data, which are then used to train the AI. This process allows the excavator to learn and replicate complex maneuvers, akin to an intern learning from a veteran. The technology promises to not only enhance efficiency but also to preserve and transfer valuable expertise in an industry facing an aging workforce and declining new entrants.

“

The barrier to entry for novice operators is lowered, and skilled professionals can work faster and more accurately.